Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

MSTOP-k implementation questions #1

Open
EtoDemerzel0427 opened this issue Jul 6, 2022 · 0 comments
Open

MSTOP-k implementation questions #1

EtoDemerzel0427 opened this issue Jul 6, 2022 · 0 comments

Comments

@EtoDemerzel0427
Copy link

Hi, I have two questions regarding the MSTOP-k implementation in ddp_final:

  1. Is this line(https://github.com/uw-mad-dash/GradCompressionUtility/blob/main/main_ddp_final.py#L506) unwanted? As you did this all_gather twice.
  2. I don't quite understand line 524, kai stores the values in tensor corresponding to indices l; after an all_gather, agg_tensor has world_size such kais, but from what I see, line 524 is just put the values in kai from different processes continuously in output tensor, disregarding their real position. What is wrong about my understanding?

Thank you in advance.
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
1 participant
@EtoDemerzel0427